A multi-condition optimization of damper with torque limiter for hybrid electric commercial vehicle based on improved grey wolf optimizer

The multi-source nature of vibration excitation and the diversity of operating modes in hybrid electric commercial vehicles result in severe drivetrain torsional vibration issues. Torsional damper is a critical solution for suppressing torsional vibration, with stiffness and stiffness angle parameters playing a decisive role in vibration reduction. However, existing passive suppression strategies are usually limited to parameter optimization under single conditions based on traditional algorithms, which not only constrain the improvement of vibration reduction performance in multi-condition operations but also struggle to address the complex nonlinear problems arising from multivariable coupling. To address these challenges, this study first designs a damper with torque limiter (DTL) with pre-damping and establishes a drivetrain torsional vibration model based on bond graph theory. Subsequently, a multi-condition optimization method for the DTL is proposed, combining a neural network mapping model and an improved grey wolf optimizer to explore the optimal matching solutions among pre-damping stiffness, primary damping stiffness, and critical stiffness angle. Finally, validation results under idle, engine start, and acceleration conditions demonstrate that the optimized DTL significantly suppresses torsional vibration. Moreover, compared to genetic algorithms, particle swarm optimization, and conventional grey wolf algorithms, the proposed method exhibits advantages of lower computational complexity, better convergence, and stronger global exploration capabilities. It is hoped that the new optimization approach proposed will inspire more innovative ideas for researchers in the field of vibration.